102
Abstract
Concentrations of antimycobacterial drugs are an intermediary link between doses admin- istered and eventual response to the drugs. Few pharmacokinetic studies have focussed on drug treatment for non-tuberculous mycobacterial (NTM) disease, although favourable treatment response occurs in just over 50% of patients despite drug treatment for at least one year. Purpose of the study: a prospective, descriptive pharmacokinetic study was per- formed to assess the plasma pharmacokinetics of rifampicin, ethambutol, clarithromycin, 14-OH-clarithromycin, azithromycin, isoniazid and moxifloxacin. Intensive pharmacokinetic sampling was performed in 14 patients with clinically relevant NTM lung disease. Pharma- cokinetic parameters were assessed and compared with available data from the literature.
Results: exposure to clarithromycin when combined with rifampicin was very low (geo-
metric mean AUC0-12h: 2.6, range 1.6-3.2 h*mg/L , Cmax: 0.3, range 0.1-0.7 mg/L ). The mean parent-to-metabolite ratios were 0.4 and 0.3 for AUC0-12h and Cmax, instead of the typical ratio of around 3, probably reflecting increased metabolism of clarithromycin to its (virtually inac- tive) 14-OH-metabolite. Exposure to rifampicin was relatively high, with all patients having a rifampicin Cmax value within the reference range. The majority of ethambutol Cmax values were within the reference range. Major conclusion: the current study re-emphasizes the relevant pharmacokinetic interaction between clarithromycin and rifampicin. This calls for a re-evaluation of the dosing strategies in NTM lung disease, as suboptimal drug exposure may contribute to inadequate response to NTM treatment.
1. Introduction
The incidence of Non-tuberculous mycobacterial (NTM) lung infections is increasing worldwide due to various factors including improved detection methods, greater awareness of clinicians,advancing age, increasing incidence of COPD, the widespread use of immunomodulating drugs and possibly an increasing environmental exposure to these bacteria.1,2 In The Netherlands, the incidence of NTM lung disease has also increased in
the past years.3 The Dutch NTM lung disease patient population differs from that in other
countries in terms of a predominance of male sex (>60% men), cavitary disease (<10%), underlying lung diseases (>75%) and the causative NTM species, among which M. avium is predominant.3-4 Successful treatment for NTM infections is hampered by misinterpretation
of drug susceptibility patterns of the causative mycobacteria, co-morbidities including underlying lung disease5 and possibly non-adherence to drug treatment due to adverse
effects.2
Unfortunately, there is limited evidence for currently recommended drug treatment regimens; these regimens are mostly based on single centre, non-comparative studies and only few prospective clinical trialshave been performed to compare possible treatment combinations.1-2
The relationship between the doses of the drugs applied and the drug concentrations achieved in patients (pharmacokinetics, PK) and the relationship between these
concentrations and response (pharmacodynamics, PD) are important knowledge gaps in NTM lung disease treatment.1,2,6 In NTM lung disease, very few studies have been performed
to study drug concentrations as an intermediary link between dose and effect.7-11
In contrast, the PK of drugs in tuberculosis (TB) treatment have been studied extensively in the past (reviewed in12-13) and it has become evident that suboptimal exposure to TB
drugs can result in inadequate response to TB treatment.14-16 In our setting, measurement
and interpretation of TB drug concentrations (Therapeutic Drug Monitoring, TDM) proved helpful in detecting low TB drug plasma concentrations as one of the possible causes of slow treatment response, treatment failure or relapse TB, and in monitoring the effect of step-wise increases in drug doses.16 Similarly, insight into the pharmacokinetics and
pharmacodynamics of drugs in NTM infections is urgently required to better comprehend how adequate response is achieved and when drug resistance and drug-related toxicity emerge.
In this study, we have determined the plasma pharmacokinetics of rifampicin, ethambutol, clarithromycin and its metabolite hydroxy-clarithromycin (OH-clarithromycin, which is far less active against M. avium complex than the parent drug), azithromycin, isoniazid and moxifloxacin in a series of patients in The Netherlands with clinically relevant NTM lung disease and we have compared the results to other series from the literature.
Materials and Methods
1 Subjects
Adult patients with M. avium complex, M. kansasii or M. malmoense pulmonary infections as diagnosed according to the American Thoracic Society (ATS) guideline,2 were eligible for the
103
104
study and were recruited from November 2010 until November 2011. Patients were either newly diagnosed or currently receiving treatment. We included patients on daily treatment with at least rifampicin and ethambutol and optionally with isoniazid, clarithromycin, azithromycin or moxifloxacin. Patients with hepatic or renal dysfunction, pregnancy, cystic fibrosis or HIV infection were excluded, as these conditions may affect PK. All patients provided written informed consent, the study was approved by the Ethical Review Board of University Medical Centre Nijmegen, The Netherlands and was registered under www. clinicaltrials.gov: NCT01051752.
2 Design and procedures
A prospective, descriptive pharmacokinetic study was performed, using the standard two- stage approach. With this approach individual pharmacokinetic parameters are estimated in the first stage. In the second stage, population characteristics of each parameter are derived by obtaining measures of central tendency and spread of the subjects’ individual parameters. The patients were treated with approved drugs supplied by our institution for at least two weeks to achieve steady state before the pharmacokinetic sampling day. Adherence in the two weeks preceding the sampling day was monitored with a medication diary in which patients recorded the time of drug intake.
Patients refrained from the intake of food from 00.00 a.m. at the PK sampling day. At 09.00 h the patients ingested all prescribed drugs on an empty stomach, under supervision of study personnel. In the case of clarithromycin, which is dosed twice daily, the second dose of 500 mg was taken 12 h later under supervision of the nurse of the hospital ward. Patients were allowed to eat, drink and take other medications starting from 12.00 am. A full pharmacokinetic curve was recorded. Serial venous blood samples were collected just prior to, and at 1.0, 1.5, 2.0, 2.5, 3.0, 4.0, 6.0, 12.0 and 24.0 h post dose, or until 12.0 h post dose for clarithromycin. Plasma was separated immediately, frozen to -80 °C and transported on dry ice for bioanalysis.
3 Bioanalysis
Total (protein bound plus unbound) plasma concentrations of rifampicin, ethambutol, clarithromycin and its metabolite 14-OH-clarithromycin, and moxifloxacin were assessed by validated high performance liquid chromatography (HPLC) methods as described before.17-19
Isoniazid was measured with liquid-liquid extraction followed by ultra performance liquid chromatography (UPLC) with ultraviolet (UV) detection. Accuracy was between 97.8% and 106.7%, dependent on the concentration level. The intra- and interassay coefficients of variation were less than 13.4 % (dependent on the concentration) over the range of 0.05 -15.1 mg/L. The lower limit of quantitation for isoniazid was 0.05 mg/L. Isoniazid containing samples were stable (<5% loss) for at least 12 months at -80oC. Azithromycin was measured
with LC-MS/MS. Accuracy was between 98.0 and 101.3%, dependent on the concentration level. The intra- and interassay coefficients of variation were less than 7.4% over the range of 0.1 mg/L to 5.0 mg/L. The lower limit of quantitation was 0.1 mg/L.
4 Pharmacokinetic analysis
Pharmacokinetic parameters were assessed with non-compartmental methods using WinNonLin version 5.0 (Pharsight Corp., Mountain View, California). The highest observed
plasma concentration was defined as Cmax, with the corresponding time as Tmax. If the concentration at 24 h post dose (C24) was below the limit of quantitation, it was calculated using the formula C24 = Clast* e (-β * (24-Tlast), in which β is the first order elimination rate
constant and Tlast is the time of the last measurable concentration (Clast). β was obtained by least squares linear regression analysis on log C versus time, with the absolute slope of the regression line being β/2.303. The area under the time-concentration curve (AUC0-24, or AUC0-12 for clarithromycin and OH-clarithromycin) was calculated using the log-linear trapezoidal rule from zero up to T=24 u or T=12 h. The apparent clearance of the drug (Cl/F) was calculated by dose/AUC0-24 or dose/AUC0-12 for clarithromycin. The volume of distribution was calculated by the equation Cl/F/ [β].
5 Statistics
Pharmacokinetic parameters were described by geometric means and range, apart from Tmax which was presented as median and range. Based on previously published reference ranges or values, Cmax values were dichotomized as either normal or low. Low concentrations were defined as <8 mg/L for rifampin,20<2 mg/L for ethambutol,20 <2.5 mg/L for clarithromycin,21
<0.3 mg/L for azithromycin22 and <3 mg/L for isoniazid20 based on standard weight-adjusted
dosing. The median peak plasma concentration for moxifloxacin at a daily dose of 400 mg as assessed in TB patients in The Netherlands was 2.5 mg/L.23
Where possible in view of patient numbers, the effects of gender, age, body weight and BMI on the PK parameters were assessed using the independent sample t-tests on the log- transformed parameters (or the Wilcoxon rank sum test for Tmax; comparison of subgroups e.g. men vs women) or Spearman’s rho on untransformed parameters (correlation of parameters). All statistical evaluations were performed with SPSS for Windows version 20 (SPSS Inc., Chicago). P values of less than 0.05 were considered significant in all analyses.